US5963106A - Double-sided pulse width modulator - Google Patents
Double-sided pulse width modulator Download PDFInfo
- Publication number
- US5963106A US5963106A US09/039,925 US3992598A US5963106A US 5963106 A US5963106 A US 5963106A US 3992598 A US3992598 A US 3992598A US 5963106 A US5963106 A US 5963106A
- Authority
- US
- United States
- Prior art keywords
- input
- output
- comparator
- bank
- amplifier
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Lifetime
Links
- 238000007599 discharging Methods 0.000 claims 3
- 239000003990 capacitor Substances 0.000 description 89
- 230000007704 transition Effects 0.000 description 30
- 230000008859 change Effects 0.000 description 6
- 238000005070 sampling Methods 0.000 description 5
- 238000009966 trimming Methods 0.000 description 5
- 230000003247 decreasing effect Effects 0.000 description 3
- 230000000630 rising effect Effects 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000000034 method Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 230000001934 delay Effects 0.000 description 1
- 230000003111 delayed effect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 238000004904 shortening Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000004513 sizing Methods 0.000 description 1
- 230000005236 sound signal Effects 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 238000005303 weighing Methods 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03K—PULSE TECHNIQUE
- H03K7/00—Modulating pulses with a continuously-variable modulating signal
- H03K7/08—Duration or width modulation ; Duty cycle modulation
Definitions
- the present invention relates to a pulse width modulated power amplifier. More particularly, the present invention relates to a Class D power amplifier where the modulation of the pulse width is preferably double-sided.
- Class D amplifiers have significant advantages compared to other types of power amplifiers such as Class A or Class B, especially with respect to power consumption.
- the output devices of a Class D amplifier operate in switched mode and, in the steady state, are either fully on or fully off.
- the fully on device has low impedance and, therefore, the power dissipation within the device itself is low.
- the fully off device is not passing current and dissipation is therefore low in this device also.
- Class A or Class B devices operate for a large proportion of time with a voltage drop across the output device while they are conducting current and therefore dissipate power during these times. As a result, Class A and B devices are less efficient and reduce the useful battery life in battery-powered equipment and/or create the need for heatsinks in high output power stages.
- Class D amplifiers include the Harris Semiconductor H1P4080 Class D amplifier described in "Class D amplifiers provide high efficiency for audio systems" Jeffrey D. Sherman, EDN May 25, 1995, pp. 103-110, and U.S. Pat. No. 4,592,087 to Killion. Both the Harris and the Killion references describe devices that accept input in analog form and operate on the principle of summing a triangular waveform together with an analog signal and comparing the result with a fixed analog voltage. The result of the comparison is used to generate the pulse width modulated output.
- Another circuit having Class D operation is the Delta-Sigma Digital to Analog Converter, in which a multi-bit digital representation is converted into a highly oversampled single bit output.
- the output is a train of pulses of equal width that, after filtering, produces the analog voltage corresponding to the multi-bit digital input.
- Delta-Sigma DAC architectures are described in "Delta-Sigma Data Converters Theory, Design and Simulation" IEEE Press, S. R. Norsworthy, R. Schreier, G. Temes Chapter 10. Operation is Class D since the output waveform is a train of digital pulses and the output devices are operated in switching mode.
- the analog filter used to filter the output pulses can be quite simple due to the high oversampling rate. In the case of audio signals, filtering can be achieved by the inductance of the loudspeaker itself.
- the oversampled converter provides a high precision output with low precision analog components at the cost of significant increase in complexity of the digital circuits.
- the circuit typically operates from a single zinc-aid cell that has a nominal voltage output of 1.3 Volts, a capacity of about 70 mAhr and, together with microphone, loudspeaker and audio processing circuits, must operate for at least 100 hours on a single cell.
- a floating point digital input having a mantissa potion and an exponent portion is input to a pulse width modulator to form a pulse width at the output that is modulated on both edges of the pulse.
- the mantissa is input to a switched capacitor amplifier in a first portion of the double-sided pulse width modulator, and the exponent is input to a capacitor bank in a second portion of the pulse width modulator.
- the product of the mantissa and the exponent form a voltage at a first input of a comparator having a discharge time that is proportional to the width of the output pulse of the double-sided modulator.
- the pulse width is modulated on both sides, and the modulation is performed in two cycles of operation.
- FIG. 1 illustrates a double-sided pulse width modulator according to the present invention.
- FIG. 2 illustrates a decode scheme for exponent bits suitable for use according to the present invention.
- FIG. 3 illustrates timing relationships between various signals in the double-sided pulse width modulator illustrated in FIG. 1 according to the present invention.
- FIG. 4 illustrates the relationship between various timing signals and selected voltage levels in the double-sided pulse width modulator illustrated in FIG. 1 according to the present invention.
- FIG. 5 illustrates a reference generator and trimming circuit suitable for use according to the present invention.
- FIGS. 6-10 illustrate alternative embodiments of a double-sided pulse width modulator according to the present invention.
- FIG. 11 illustrates an alternative decode scheme for exponent bits suitable for use according to the present invention.
- a double-sided pulse width modulator 10 according to the present invention is illustrated. It is contemplated according to the present invention, that the double-sided pulse width modulator 10 will operate on a digital input that in the preferred embodiment is a signed floating point number which includes 1 bit, S, for the sign, 8 bits, m0-m7, for the mantissa, and 3 bits, e0-e2, for the exponent. In alternative embodiments to be described below, other number representations and decoding schemes in accordance with the principles of the present invention will be disclosed.
- the sampling of the digital input is uniform, and the output, U, of the double-sided pulse width modulator 10 is a pulse wave form of fixed frequency.
- the width of the output pulse is modulated in proportion to the digital input value.
- the modulation of the output pulse is double-sided, rather than single sided.
- the width of the pulse will be increased or decreased on both the rising edge and the falling edge of the pulse.
- only one of the edges of the pulse is increased or decreased while the remaining edge is kept fixed relative to the sampling interval.
- the double-sided pulse width modulator 10 includes a switched capacitor amplifier 12 having a non-inverting input, an inverting input and an output, and a comparator 14 having a non-inverting input, an inverting input and an output.
- the double-sided pulse width modulator 10 according to the present invention is very tolerant of many of the sources of non ideal behavior. This is due to the fact that many of the sources of non ideal behavior have a similar effect on both the positive and negative edges of the pulse width output, U, of the double-sided pulse-width modulator 10.
- Some examples of these sources of non ideal behavior include the switched capacitor amplifier 12 and comparator 14 offsets, CLK feed through, inaccuracies in Vdd, and well-known propagation delays in logic gates and comparator 14.
- the non-inverting input of the switched capacitor amplifier 12 is connected to a reference potential Vdd/2, and the inverting input of the switched capacitor amplifier 12 is connected to one side of each capacitor in a first bank of capacitors 16-1 through 16-8.
- the other side of each individual capacitor in the first bank of capacitors 16-1 through 16-8 is connected, respectively, to the output of one of the two-input NAND gates 18-1 through 18-8.
- the first bank of capacitors 16-1 through 16-8 are weighted as C, 2 C, 4 C, 8 C, 16 C, 32 C, 64 C, and 128 C.
- the two-input NAND gates 18-1 through 18-8 driving the first bank of capacitors 16-1 through 16-8 are illustrated as logic gates. It should be appreciated by those of ordinary skill in the art that the two-input NAND gates 18-1 through 18-8 may be replaced with transistor switches
- a first switch 20, and a capacitor 22 weighted at 640 C are connected between the inverting input and the output of switched capacitor amplifier 12.
- First switch 20 is controlled by a first timing signal, ⁇ 1, to be described below.
- the capacitor 22 in the feedback path of the switched capacitor amplifier 12 is sized to keep the output of the switched capacitor amplifier 12 within the limits of the voltage supplied to the switched capacitor amplifier 12 so that the switched capacitor amplifier 12 does not saturate.
- a first input of each of the two-put NAND gates 18-1 through 18-8 is input with one of the mantissa bits m0-m7.
- a second input of each of the two-input NAND gates 18-1 through 18-8 is connected to the output of a first two-input XNOR gate 24.
- a first input of the first two-input XNOR gate 24 is input with a second timing signal, ⁇ 2, to be explained below, and a second input of the first two-input XNOR gate 24 is connected to the output of a second two-input XNOR gate 26.
- a first input of the second two-input XNOR gate 26 is input with the sign bit, S, and a second input of the second two-input XNOR gate 26 is input with the output, U, of the double-sided pulse width modulator 10.
- the non-inverting input of the comparator 14 is connected to a reference potential of 0.1 Vdd, and the inverting input of the comparator 14 is connected to one side of each capacitor in a second bank of capacitors 28-1 through 28-8.
- the enable signal for comparator 14 is connected to a third timing signal, 43, to be described below.
- the second bank of capacitors 28-1 through 24-8 are weighted as C', C', 2 C', 4 C', 8 C', 16 C', 32 C', and 64 C', respectively.
- capacitor 28-1 is connected directly to the output of amplifier 12, while the other side of each individual capacitor in the second bank of capacitors 28-2 through 28-8 may be connected alternatively by a separate pair of switches for each of the individual capacitors 28-2 through 28-8 to either the output of amplifier 12 or to ground.
- the second bank of capacitors 28-1 through 28-8 may be implemented with the desired number of unit sized capacitors C', or may be implemented according to other practices employed to ensure good matching as is known to those of ordinary skill in the art. Although the capacitors within each of the first and second banks of capacitors 16-1 through 16-8, and 28-1 and 28-8 must be matched within their respective banks, it is not required that the capacitors between the first and second banks of capacitors 16-1 through 16-8, and 28-1 and 28-8 be matched.
- each pair of switches 30-1a and 30-1b through 30-7a and 30-7b a first terminal in each of the switches in each pair of switches is connected to one of the capacitors in the second bank of capacitors 28-2 through 28-8.
- a second terminal of the switches 30-1a through 30-7a is connected to the output of amplifier 12, and a second terminal of the switches 30-1b through 30-7b is connected to ground.
- the pairs of switches 30-1a and 30-1b through 30-7a and 30-7b may be implemented in a variety of ways known to those of ordinary skill in the art, including, for example, pairs of pass transistors.
- Each of the switches in the pairs of switches 30-1a and 30-1b through 30-7a and 30-7b has a control input connected to the output of decode logic 32.
- the exponent bits e0-e2 are decoded by decode logic 32 to provide eight pairs of bits, E1/E1 through E7/E7 as control signals for the eight pairs of switches 30-1a and 30-1b through 30-7a and 30-7b.
- the implementation of the decode logic 32 is well within the level of skill of those of ordinary skill in the art and will not be disclosed herein to avoid over complicating the disclosure and thereby obscuring the present invention.
- the exponent bits e0-e2 are decoded into switch control bits E1/E1 through E7/E7 according to the decode scheme indicated in FIG. 2.
- a decode bit is a logic ⁇ 1 ⁇
- the associated switch 30-1a through 30-7a is turned on and the ⁇ voltage at the output of the switch capacitor amplifier 12 is connected to the capacitor 28-1 through 28-8 associated with the switch 30-1a through 30-17a.
- the decode bit is a logic ⁇ 0 ⁇
- the switch 30-1a through 30-7a is turned off and the complemented switch 30-1b through 30-7b connects the associated capacitor 28-1 through 28-8 to ground.
- the combination of the input decoding and the capacitor sizing in the bank of second capacitors 28-1 through 28-8 implements an exponent for the mantissa connected to the NAND gates 18-1 through 18-8.
- N-channel MOS transistors 34, 36, and 38 are also connected to the inverting input of comparator 14.
- the source of N-channel MOS transistor 34 is connected to the drain of N-channel MOS transistor 36
- the source of N-channel MOS transistor 36 is connected to the drain of N-channel MOS transistor 38
- the source of N-channel MOS transistor 38 is connected to ground.
- N-channel MOS transistors 36, 38, and 40 are controlled by signals Vbias1, Vbias2, and a third timing signal ⁇ 3, respectively, to be described below.
- the N-channel MOS transistors 34, 36, and 38 form a current source.
- a second switch 40 is controlled by the first timing signal, ⁇ 1, and is connected between the output of amplifier 12 and the inverting input of comparator 14.
- a third portion of the double-sided pulse width modulator 10 shown within dashed line 42 is responsive to the output of the comparator 14 and the clock signal, CLK, to provide first, second and third timing signals ⁇ 1, ⁇ 2, and ⁇ 3 and the output U, of the double-sided pulse width modulator 10.
- Portion 42 includes first, second and third D-type flip-flops, 44, 46, and 48, buffer circuits 50 and 52, first and second two-input OR gates 54 and 56, and inverter 58.
- Buffer circuit 50 may be implemented as a bootstrap circuit to generate a HIGH level on timing signal ⁇ 1 that is greater than Vdd, as is well known in the art.
- the boosted voltages on timing signal ⁇ 1 are selected to improve the conduction of switches 20 and 40 for all possible voltages of analog signals applied to switches 20 and 40, especially for low values of Vdd.
- double-sided pulse width module 10 a voltage dependent on the values of the mantissa and exponent bits, m0-m7 and e0-e2, respectively, is placed at the inverting input of the comparator 14, according to timing relationships to be described.
- the voltage at the inverting input of the comparator 14 is then discharged through N-Channel MOS transistors 34, 36 and 38.
- the input of comparator 14 discharges to the reference voltage, preferably 0.1 Vdd at the non-inverting input of the comparator 14, a HIGH level is output from the comparator 14.
- the width of the output pulse of the double-sided output pulse modulator 10 is proportional to the length of the time of the discharge of the voltage at the inverting input of comparator 14.
- the voltage placed at the inverting input of the comparator 14 falls to the reference voltage, preferably, 0.1 Vdd at the non-inverting input of the comparator 14, a step is output from the comparator 14.
- the present invention is tolerant to non-linearities that add and subtract equally to the discharge times. For example, it will be appreciated by those of ordinary skill in the art, that as the voltage at the node connected to the inverting input of the comparator 14 approaches the reference voltage, 0.1 Vdd, that the cascode effect diminishes, and a reduced drain voltage on the N-Channel MOS transistor 36 may cause a reduction in the drain current.
- the portion 42 of the double-sided output pulse modulator 10 has two inputs, the CLK input and the output of comparator 14.
- the CLK input is connected to the first input of the first two-input OR gate 54, a first input of the second two-input OR gate 56, and through the inverter 58 to the clock input of second D-type flip-flop 46.
- the output of the comparator 14 is connected to the second input of the first two-input OR gate 54.
- the relationship between the CLK signal and the timing signals ⁇ 1, ⁇ 2, and ⁇ 3, can be understood upon inspection of circuit portion 42, and by reference to the timing diagrams in FIG. 3.
- the output, Q2 of second D-type flip-flop 46 makes a transition from LOW to HIGH.
- the output, Q2 of second D-type flip-flop 46 which is connected to the reset input of first D-type flip-flop 44, in going from LOW to HIGH causes the output, Q1, of first D-type flip-flop 44 to make a transition from HIGH to LOW, and also for the inverted output, Q1B of first D-type flip-flop 44 to make a transition from LOW to HIGH.
- the inverted output, Q1B, of second D-type flip-flop 44 which is connected to the reset input of second D-type flip-flop 46, in making a transition from LOW to HIGH causes the second D-type flip-flop 46 to be reset, therefore, causing the output, Q2, of second D-type flip-flop 46 to make a transition from HIGH to LOW.
- the output of OR gate 56 is also LOW.
- the output ⁇ 1 of first buffer circuit 50 whose input is connected to the output of OR gate 56, makes a transition from HIGH to LOW.
- the input of second buffer circuit 52 is also connected to the output of OR gate 56, accordingly the outputs, 42 and 43, of second buffer circuit 52 both make a transition from LOW to HIGH.
- the switches 20 and 40 in FIG. 1 are turned on, and the node at the inverting input of the comparator 14, and the node at the output of the switch capacitor amplifier 12 are both charged to Vdd/2.
- the CLK signal makes a transition from HIGH to LOW
- the ⁇ 1 signal as described above, also makes a transition from HIGH to LOW.
- the ⁇ 2 signal whose phase is inverse to the ⁇ 1 signal, then makes a transition from LOW to HIGH.
- the ⁇ 2 signal is applied to a first input of exclusive XNOR gate 24.
- the output of the XNOR gate 24 is determined by the inputs S and U to the XNOR gate 26. This is an important feature of the present invention because the state of the bit S and the state of the output U are used to determine the polarity of the voltage transition into the first capacitor array 16-1 through 16-8. A change in voltage occurs at the output of the switched capacitor amplifier 12 due to the voltage change in the first bank of capacitors 16-1 through 16-8 connected to the inverting input of the switched capacitor amplifier 12.
- Whether a capacitor in the first bank of capacitors 16-1 through 16-8 receives a logic transition to place charge on the capacitor depends upon the value of the mantissa bit supplied to the NAND gate 18-1 through 18-8 connected to the corresponding capacitor 16-1 through 16-8 respectively.
- the output of the NAND gate 18-1 through 18-8 remains HIGH, and when the mantissa bit being fed to the NAND gate 18-1 through 18-8 is a HIGH logic level, the complement of the output of the XNOR gate 24 is passed through the NAND gate 18-1 through 18-8 to the capacitors in the first bank of capacitors 16-1 through 16-8 associated with the NAND gates 18-1 through 18-8 having a HIGH logic level from the mantissa bit.
- the voltage supplied to the first bank of capacitors 16-1 through 16-8 is related to the voltage supplied to the NAND gates 18-1 through 18-8. It should be appreciated that the voltage supplied to NAND gates 18-1 through 18-8 may be different than the voltage supplied to the remainder of the double-sided pulse width modulator 10, as indicated in FIG. 1 by the input signal VC. A change in the voltage to the NAND gates 18-1 through 18-8, will result in a change in the voltage output from the switched capacitor amplifier 12, and the amount of voltage stored at the inverting input of the comparator 14.
- the voltage supplied to NAND gates 18-1 through 18-8 may be varied, for example, as a volume control to adjust the output.
- a HIGH logic level out of the XNOR gate 24 places a LOW logic level at the outputs of the NAND gate 18-1 through 18-8 having a HIGH logic level at their mantissa bit. This increases the voltage at the inverting input into the comparator 14, which in turn increases the discharge time of the node at the inverting input to comparator 14. It should be appreciated that the discharge time of the node of the inverting input comparator 14 controls the modulation of the pulse width output, U. Accordingly, an increase in discharge time increases the pulse width of the output.
- the first transition at the output of the switched capacitor amplifier 12 is shown as a voltage of ⁇ above Vdd/2.
- the A voltage is calculated as follows:
- the ⁇ voltage is connected to selected ones of the second bank of capacitors 28-1 through 28-8 by the switches 30-1a through 30-7a by the outputs of the decode logic 32 according to the exponent value.
- a voltage is placed at the inverting input of the comparator 14 that is equal to ⁇ times the value of the exponent, divided by 128.
- the voltage at the inverting input at the comparator 14 is then discharged by N-Channel MOS transistors 34, 36 and 38.
- the output of comparator 14 makes a transition from a LOW state to a HIGH state.
- the feedback capacitor 22 of the switched capacitor amplifier 12 may be sized greater than 640 C in order to reduce the ⁇ voltage at the output of the switched capacitor amplifier 12, as described above, and to match the starting voltage found at the node connected to the inverting input of the comparator 14 to the linear region of the current source formed N-Channel MOS transistors 34, 36 and 38.
- the reference voltages for the comparator 14 and the switched capacitor amplifier 12 are 0.1 Vdd and Vdd/2, respectively.
- reference voltages were selected to maximize the voltage swing on the node connected to the inverting input of the comparator 14 without a loss of response from the comparator 14, without a loss of linearity in the current source formed by N-Channel MOS transistors 34, 36 and 38 and without saturation of the switched capacitor amplifier 12.
- the reference voltages Vdd/2 and 0.1 Vdd, and gate voltage Vbias are generated by the reference generator and current trim circuit 70 illustrated in FIG. 5.
- the reference voltages VDD/2 and 0.1 VDD are generated by first second, and third reference generating resistors 72, 74 and 76 connected in series between VDD and ground.
- the first second, and third reference generating resistors 72, 74 and 76 respectively, have a total resistance equal to R1, and individual resistances equal to 0.5 R1, 0.4 R1, and 0.1 R1, respectively.
- a first end of first reference generating resistor 72 is connected to VDD and a second end of first reference generating resistor 72 is connected to the inverting input of an amplifier 78, and also to a first end of second reference generating resistor 74. Accordingly, the voltage formed at the inverting input of amplifier 78 is equal to VDD/2.
- a second end of the second reference generating resistor 74 is connected to a first end of third reference generating resistor 76, and the second end of third reference generating resistor 76 is connected to ground.
- the voltage formed at the connection between the second end of second reference generating resistor 74 and the first end of third reference generating resistor 76 is 0.1 VDD.
- a current mirror is formed from P-Channel MOS transistors 80 and 82, and trimming network is formed by trim resistor 84 and variable trim resistor 86.
- the sources and gates of the P-Channel MOS transistors 80 and 82 are connected to VDD, and the output of amplifier 78, respectively.
- the drain of P channel MOS transistor 80 is connected to the non-inverting input of amplifier 78, and to a first end of the trim resistor 84.
- trim resistor 84 is connected to a first end of a variable trim resistor 86 formed by resistor and N-Channel transistor pairs 88-0a and 88-0b through 88-ma and 88-mb.
- the second end of the variable trim resistor 86 is connected to ground.
- Digital trim bits Tr-0 through Tr-m are input to the N-Channel MOS transistors 88-0b through 88-mb to adjust the value of the variable trim resistor 86.
- the resistors 88-0a through 88-ma are weighted in binary to correspond to the binary weights associated with the trim bits Tr-0 through Tr-m.
- the trim bits Tr-0 through Tr-m may be provided by a non-volatile memory such as an EEPROM, and therefore may be changed to adjust the value of the variable trim resistor 86.
- first, second and third N-Channel MOS transistors 90, 92 and 94 Connected in series to the drain of P-Channel MOS transistor 82 are first, second and third N-Channel MOS transistors 90, 92 and 94.
- the drain of N-channel MOS transistor 90 is connected to the drain of P-channel MOS transistor 82 and also to the gate of N-channel MOS transistor 92.
- the source of N-channel MOS transistor 90 is connected to the drain of N-channel MOS transistor 92.
- the drain of N-channel MOS transistor 94 is connected to the source of N-channel MOS transistor 92, and the source of N-channel MOS transistor 94 is connected to ground.
- the gate of N-channel MOS transistor is connected to a DC bias, Vbias1, that has been generated by an external source, and which is also connected to the gate of the N-channel MOS transistor 34 illustrated in FIG. 1.
- Vbias1 a DC bias
- the gate of N-channel MOS transistor 94 is connected to VDD.
- the current flowing through first, second and third N-Channel MOS transistors 34, 36 and 38 determines the discharge time of the voltage at the inverting input of comparative 14. It is preferable that the voltage at the inverting input of comparator 14 be discharged so that the comparator 14 will trigger prior to each pulse of the CLK signal. Accordingly, for all mantissa and exponent values, the discharge current through N-channel MOS transistors 34, 36 and 38, respectively, is designed to ensure that the voltage at the input to comparator 14 is discharged to 0.1 VDD.
- the value of the discharge current can also be adjusted.
- the current trimming provided for the Vbias2 signal can be employed to compensate for a variation in the values of the resistor 84, resistors 88-0a trough 88-ma used to generate the discharge current, and also for the unit capacitor values in the second bank of capacitors 28-1 through 28-8. Trimming to compensate for these, and other variations due to manufacturing tolerances helps to provide high precision output pulse widths. Also, where a loudspeaker is connected to the output of the double-sided pulse width modulator 10, trimming by the Vbias2 signal can match the output pulse width to the characteristics of the loud speaker.
- the trim settings for a particular circuit may be modified to adjust the output amplitude as given by the pulse duty cycle at output U.
- the output amplitude can be trimmed to compensate for variations in individual loudspeaker characteristics, as may occur due to manufacturing variations. It will also be appreciated by those of ordinary skill in the art that since the Vbias2 signal is proportional to Vdd, any variations in the ⁇ voltage are compensated for because the ⁇ voltage is also proportional to Vdd.
- the output, Q1 of the first D type flip-flop 44 follows.
- the Q1 output through OR gate 56 results in the ⁇ 1 signal making a transition from a LOW state to a HIGH state, the ⁇ 2 signal making a transition from a HIGH state to a LOW state, the ⁇ 3 signal making a transition from a HIGH state to a LOW state, and the output, U, of third D-type flip-flop 48 making a change of state.
- the ⁇ 3 signal makes the transition from HIGH to LOW that the comparator 14 output which is controlled by the 43 signal also makes a transition from HIGH to LOW.
- the output of the switched capacitor amplifier 12 and the inverting input of the comparator 14 are again set to Vdd/2.
- this causes the ⁇ 1 signal to make a transition from a LOW state to a HIGH state, and the ⁇ 2 and ⁇ 3 signals to make a transition from a HIGH state to a LOW state.
- the output of the third D-type flip-flop 44 now makes the transition from a LOW state to a HIGH state. From the two clock cycles just described, it will be appreciated by those of ordinary skill in the art that the falling edge of the output, U, was initiated by the first clock pulse, and that the rising edge of the output, U, was initiated by the second clock pulse.
- the pulse width is modulated on both sides, and that in two cycles of operation and two clock pulses the modulation is performed.
- the dashed lines in the waveforms of FIG. 4 correspond to a 50% duty cycle and a net zero output level the bits m0-m7 are 00000000.
- the next cycle consider the value of the output, U, being at a LOW logic state and the sign bit, S, still being HIGH.
- the positive edge of the signal ⁇ 2 creates a LOW signal into the first bank of capacitors 16-1 through 16-8 to lengthen the LOW state of the output, U, thereby shortening the next successive pulse width.
- FIGS. 6, 7, 8, 9, and 10 For those portions of FIGS. 6-10, that are the same as the embodiment illustrated in FIG. 1 the reference numerals have been retained.
- FIG. 6 in a first alternative embodiment of double sided pulse-width modulator 98, the second bank of capacitors 28-1 through 28-8, and the pairs of switches 30-1a and 30-1b through 30-7a and 30-7b depicted in FIG. 1 are replaced by a bank of capacitors 100-1 through 100-8, and the pairs of switches 102-0a and 102-0b through 102-7a and 102-7b.
- one side of each of the capacitors in the second bank of capacitors 100-1 through 100-8 is connected to the inverting input node of comparator 14, and a second side of each of the capacitors 100-1 through 100-8 is connected to a pair of switches 102-0a, and 102-0b through 102-7a and 102-7b.
- the capacitor values for the capacitors 100-1 through 100-8 in the second bank of capacitors are C', 2 C', 4 C', 8 C', 16 C', 32 C', 64 C', and 128 C'.
- FIG. 11 The decoding scheme of the decode logic 32 for the pairs of switches 102-0a, and 102-0b through 102-7a and 102-7b for this alternative embodiment is depicted in FIG. 11. From the decoding scheme it can be observed that only one of the switches 102-0a through 102-7a is conducting at any one time. The remaining elements in FIG. 6 operate as described above with regard to FIG. 1.
- FIG. 7 in a second alternative embodiment of double sided pulse-width modulator 108, the decode logic 32 illustrated in FIG. 6 has been eliminated, and a second set of binary inputs n0-n7 and their compliments, control the pairs of switches 110-0a and 110-0b through 110-7a and 110-7b. Accordingly, in the embodiment depicted in FIG. 7, the output pulse width of the double-sided modulator 10 will be proportional to the product of the mantissa bits m0-m7, and the binary word n0-n7.
- the second bank of capacitors 28-1 through 28-8, and the pairs of switches 30-1a and 30-1b through 30-7a and 30-7b depicted in FIG. 1 are replaced by a single capacitor 120 having a value of C'.
- the digital input consists of the mantissa bits m0-m7, and the sign bit S.
- the sign bit S and the XNOR gate 26 are eliminated, and the output U is connected to the second input of XNOR gate 24. Accordingly, the pulse out width, U, is a function of the mantissa bits m0-m7.
- FIG. 9 in a fourth alternative embodiment of double sided pulse-width modulator 128, the second bank of capacitors 28-1 through 28-8, and the pairs of switches 30-1a and 30-1b through 30-7a and 30-7b depicted in FIG. 1 are replaced with a single capacitor 130 and an array of switched current devices 132.
- each transistor in a first set of N-channel MOS transistor 134-1 through 134-8 has a drain connected to the inverting input of comparator 14 and a source connected to a drain of one of the transistors in a second set of N-channel MOS transistors 136-1 through 136-8.
- Each of the sources in the second set of N-channel MOS transistors 136-1 through 136-8 is connected to a drain of one of the transistors in a third set of N-channel MOS transistors 138-1 through 138-8.
- the source of each of the transistors in the third set of N-channel MOS transistors 138-1 through 138-8 is connected to the drain of an N-channel MOS transistor 140.
- the source of N-channel MOS transistor 140 is connected to ground.
- the gates of each of transistors in the first set of N-channel MOS transistors 134-1 through 134-8 is connected to VBIAS1.
- the gates of the transistors in the second set of N-channel MOS transistors 136-1 through 136-8 are connected to VBIAS2.
- the gates of the transistors in the third set of N-channel MOS transistors 138-1 through 138-8 are connected to the outputs E0-E7 of the decode logic 32, respectively.
- the gate of N-channel MOS transistor 140 is connected to timing signal ⁇ 3 as described above.
- the capacitor 130 and the array of switched current devices 132 provide a fixed capacitor and a variable current sink.
- the first, second and third sets of N-channel MOS transistors 134-1 through 134-8, 136-1 through 136-8 and 138-1 through 138-8 are scaled so that the current passing through each string of three series transistors, 134, 136, and 138 provides a binary weighing assigned to the decode signals E0-E7 from the decode logic 32.
- a switch 150 that is controlled by the first timing signal ⁇ 1 is disposed between the inverting and non-inverting input of comparator 14.
- the non-inverting input is connected to a Vref so that when the first timing signal ⁇ 1 is high, the inverting input of comparator 14 is initialized to Vref.
- the value of Vref is preferably Vdd/2 to provide maximum output amplitude.
- the capacitor 152 is scaled to be greater than 128 C' so that when the voltage VC described above is less than or equal to Vdd, the voltage at the inverting input of comparator 14 will be greater than the voltage V ref placed at the inverting input of comparator 14 by the first timing signal ⁇ 1.
- the discharge time at the inverting input of comparator 14 is either increased or decreased by an amount related to the product of the mantissa bits m0-m7 and the exponent values from the decode logic 32 to modulate the output, U.
- single-sided modulation may be obtained from the embodiments described herein with only slight modification.
- single-sided modulation may be obtained by positioning the output pulse edges of a first transition direction at a constant time delay with respect to the input clock, and by positioning, as described in the embodiments set forth above, the output pulse edges of a second transition direction at a variable time delay with respect to the input clock.
Abstract
Description
Δ=(Value of the mantissa×Vdd)/640.
Claims (3)
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US09/039,925 US5963106A (en) | 1998-03-16 | 1998-03-16 | Double-sided pulse width modulator |
PCT/US1999/005699 WO1999048205A1 (en) | 1998-03-16 | 1999-03-16 | Double-sided pulse width modulator |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US09/039,925 US5963106A (en) | 1998-03-16 | 1998-03-16 | Double-sided pulse width modulator |
Publications (1)
Publication Number | Publication Date |
---|---|
US5963106A true US5963106A (en) | 1999-10-05 |
Family
ID=21908097
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US09/039,925 Expired - Lifetime US5963106A (en) | 1998-03-16 | 1998-03-16 | Double-sided pulse width modulator |
Country Status (2)
Country | Link |
---|---|
US (1) | US5963106A (en) |
WO (1) | WO1999048205A1 (en) |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6072534A (en) * | 1996-12-05 | 2000-06-06 | Sgs-Thomson Microelectronics S.A. | Device for the subdivision of the period of a signal into N quasi-equal parts |
US6163287A (en) * | 1999-04-05 | 2000-12-19 | Sonic Innovations, Inc. | Hybrid low-pass sigma-delta modulator |
US6408318B1 (en) | 1999-04-05 | 2002-06-18 | Xiaoling Fang | Multiple stage decimation filter |
US6445321B2 (en) | 1999-04-05 | 2002-09-03 | Sonic Innovations, Inc. | Hybrid low-pass sigma-delta modulator |
US20030179831A1 (en) * | 2002-03-21 | 2003-09-25 | Deepnarayan Gupta | Power amplifier linearization |
US6670837B2 (en) * | 2001-09-26 | 2003-12-30 | Tempo Research Corporation | Time domain reflectometer with digitally generated variable width pulse output |
US6768779B1 (en) * | 1997-04-02 | 2004-07-27 | Bang & Olufsen Powerhouse A/S | Pulse referenced control method for enhanced power amplification of a pulse modulated |
US6922134B1 (en) * | 1998-04-14 | 2005-07-26 | The Goodyear Tire Rubber Company | Programmable trimmer for transponder |
US20100097153A1 (en) * | 2006-10-27 | 2010-04-22 | Leonard Rexberg | Switched Modulation of a Radio-Frequency Amplifier |
US9559642B2 (en) | 2015-01-02 | 2017-01-31 | Logitech Europe, S.A. | Audio delivery system having an improved efficiency and extended operation time between recharges or battery replacements |
US9906191B1 (en) | 2010-08-02 | 2018-02-27 | Hypres, Inc. | Superconducting multi-bit digital mixer |
Citations (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO1983000221A1 (en) * | 1981-07-09 | 1983-01-20 | Riesenberg, Klaus | Device for measuring the fuel consumption of an internal combustion engine |
US4545065A (en) * | 1982-04-28 | 1985-10-01 | Xsi General Partnership | Extrema coding signal processing method and apparatus |
US4588960A (en) * | 1982-02-01 | 1986-05-13 | University Of Toronto Innovations Foundation | Output circuits of class B type electronic amplifiers |
US4592087A (en) * | 1983-12-08 | 1986-05-27 | Industrial Research Products, Inc. | Class D hearing aid amplifier |
US4685042A (en) * | 1984-07-03 | 1987-08-04 | Unitron, Inc. | Modulator control for inverter |
US4689819A (en) * | 1983-12-08 | 1987-08-25 | Industrial Research Products, Inc. | Class D hearing aid amplifier |
US5014016A (en) * | 1989-04-13 | 1991-05-07 | Beltone Electronics Corporation | Switching amplifier |
US5099856A (en) * | 1989-11-08 | 1992-03-31 | Etymotic Research, Inc. | Electrode isolation amplifier |
US5162751A (en) * | 1990-09-06 | 1992-11-10 | U.S. Philips Corporation | Amplifier arrangement |
EP0534804A1 (en) * | 1991-09-27 | 1993-03-31 | Exar Corporation | Class-D BICMOS hearing aid output amplifier |
EP0590903A1 (en) * | 1992-09-30 | 1994-04-06 | Exar Corporation | Output limiter for class-D BICMOS hearing aid output amplifier |
US5317640A (en) * | 1990-10-30 | 1994-05-31 | Ascom Audiosys Ag | Switched amplifier |
US5378933A (en) * | 1992-03-31 | 1995-01-03 | Siemens Audiologische Technik Gmbh | Circuit arrangement having a switching amplifier |
US5451893A (en) * | 1994-05-13 | 1995-09-19 | Samsung Semiconductor, Inc. | Programmable duty cycle converter |
US5512857A (en) * | 1994-11-22 | 1996-04-30 | Resound Corporation | Class AB amplifier allowing quiescent current and gain to be set independently |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR0154797B1 (en) * | 1995-07-27 | 1998-12-15 | 김광호 | Pulse width modulation control circuit |
-
1998
- 1998-03-16 US US09/039,925 patent/US5963106A/en not_active Expired - Lifetime
-
1999
- 1999-03-16 WO PCT/US1999/005699 patent/WO1999048205A1/en active Application Filing
Patent Citations (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO1983000221A1 (en) * | 1981-07-09 | 1983-01-20 | Riesenberg, Klaus | Device for measuring the fuel consumption of an internal combustion engine |
US4588960A (en) * | 1982-02-01 | 1986-05-13 | University Of Toronto Innovations Foundation | Output circuits of class B type electronic amplifiers |
US4545065A (en) * | 1982-04-28 | 1985-10-01 | Xsi General Partnership | Extrema coding signal processing method and apparatus |
US4592087A (en) * | 1983-12-08 | 1986-05-27 | Industrial Research Products, Inc. | Class D hearing aid amplifier |
US4689819A (en) * | 1983-12-08 | 1987-08-25 | Industrial Research Products, Inc. | Class D hearing aid amplifier |
US4592087B1 (en) * | 1983-12-08 | 1996-08-13 | Knowles Electronics Inc | Class D hearing aid amplifier |
US4689819B1 (en) * | 1983-12-08 | 1996-08-13 | Knowles Electronics Inc | Class D hearing aid amplifier |
US4685042A (en) * | 1984-07-03 | 1987-08-04 | Unitron, Inc. | Modulator control for inverter |
US5014016A (en) * | 1989-04-13 | 1991-05-07 | Beltone Electronics Corporation | Switching amplifier |
US5099856A (en) * | 1989-11-08 | 1992-03-31 | Etymotic Research, Inc. | Electrode isolation amplifier |
US5162751A (en) * | 1990-09-06 | 1992-11-10 | U.S. Philips Corporation | Amplifier arrangement |
US5317640A (en) * | 1990-10-30 | 1994-05-31 | Ascom Audiosys Ag | Switched amplifier |
US5247581A (en) * | 1991-09-27 | 1993-09-21 | Exar Corporation | Class-d bicmos hearing aid output amplifier |
US5389829A (en) * | 1991-09-27 | 1995-02-14 | Exar Corporation | Output limiter for class-D BICMOS hearing aid output amplifier |
US5506532A (en) * | 1991-09-27 | 1996-04-09 | Exar Corporation | Output limiter for class-D BICMOS hearing aid output amplifier |
EP0534804A1 (en) * | 1991-09-27 | 1993-03-31 | Exar Corporation | Class-D BICMOS hearing aid output amplifier |
US5378933A (en) * | 1992-03-31 | 1995-01-03 | Siemens Audiologische Technik Gmbh | Circuit arrangement having a switching amplifier |
EP0590903A1 (en) * | 1992-09-30 | 1994-04-06 | Exar Corporation | Output limiter for class-D BICMOS hearing aid output amplifier |
US5451893A (en) * | 1994-05-13 | 1995-09-19 | Samsung Semiconductor, Inc. | Programmable duty cycle converter |
US5512857A (en) * | 1994-11-22 | 1996-04-30 | Resound Corporation | Class AB amplifier allowing quiescent current and gain to be set independently |
Cited By (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6072534A (en) * | 1996-12-05 | 2000-06-06 | Sgs-Thomson Microelectronics S.A. | Device for the subdivision of the period of a signal into N quasi-equal parts |
US6768779B1 (en) * | 1997-04-02 | 2004-07-27 | Bang & Olufsen Powerhouse A/S | Pulse referenced control method for enhanced power amplification of a pulse modulated |
US6922134B1 (en) * | 1998-04-14 | 2005-07-26 | The Goodyear Tire Rubber Company | Programmable trimmer for transponder |
US6466146B1 (en) | 1999-04-05 | 2002-10-15 | Sonic Innovations, Inc. | Hybrid low-pass sigma-delta modulator |
US6445321B2 (en) | 1999-04-05 | 2002-09-03 | Sonic Innovations, Inc. | Hybrid low-pass sigma-delta modulator |
US6408318B1 (en) | 1999-04-05 | 2002-06-18 | Xiaoling Fang | Multiple stage decimation filter |
US6163287A (en) * | 1999-04-05 | 2000-12-19 | Sonic Innovations, Inc. | Hybrid low-pass sigma-delta modulator |
US6670837B2 (en) * | 2001-09-26 | 2003-12-30 | Tempo Research Corporation | Time domain reflectometer with digitally generated variable width pulse output |
US20030179831A1 (en) * | 2002-03-21 | 2003-09-25 | Deepnarayan Gupta | Power amplifier linearization |
US7313199B2 (en) * | 2002-03-21 | 2007-12-25 | Hypres, Inc. | Power amplifier linearization |
US20100097153A1 (en) * | 2006-10-27 | 2010-04-22 | Leonard Rexberg | Switched Modulation of a Radio-Frequency Amplifier |
US7932790B2 (en) * | 2006-10-27 | 2011-04-26 | Telefonaktiebolaget Lm Ericsson (Publ) | Switched modulation of a radio-frequency amplifier |
US9906191B1 (en) | 2010-08-02 | 2018-02-27 | Hypres, Inc. | Superconducting multi-bit digital mixer |
US9559642B2 (en) | 2015-01-02 | 2017-01-31 | Logitech Europe, S.A. | Audio delivery system having an improved efficiency and extended operation time between recharges or battery replacements |
US9722544B2 (en) | 2015-01-02 | 2017-08-01 | Logitech Europe S.A. | Audio system having an improved efficiency and extended operation time |
US9742357B2 (en) | 2015-01-02 | 2017-08-22 | Logitech Europe S.A. | Audio system having an improved efficiency and extended operation time |
Also Published As
Publication number | Publication date |
---|---|
WO1999048205A1 (en) | 1999-09-23 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US5506532A (en) | Output limiter for class-D BICMOS hearing aid output amplifier | |
KR101118867B1 (en) | Power amplifying circuit, dc-dc converter, peak holding circuit, and output voltage control circuit including the peak holding circuit | |
US5614864A (en) | Single-ended to differential converter with relaxed common-mode input requirements | |
US7330069B2 (en) | Digital switching amplifier | |
US7414471B2 (en) | Common-mode technique for a digital I/P class D loop | |
US9876489B1 (en) | Method of implementing a differential integrating phase interpolator | |
JP3270081B2 (en) | High efficiency power audio amplifier including two amplifiers with a single source | |
US7679435B2 (en) | Digital input class-D amplifier | |
US5963106A (en) | Double-sided pulse width modulator | |
KR20020027258A (en) | Power supply pulse width modulation(pwm) control system | |
US6917242B2 (en) | Power amplifier | |
GB2610917A (en) | Chopped triangular wave PWM quantizer and PWM modulator having quantizer with controllable analog gain and calibratable for multi-non-ideal gain-affecting | |
JP4461813B2 (en) | Pulse width modulation amplifier | |
US6734725B2 (en) | Power amplifier | |
CN112886933B (en) | Class D audio amplifier, adaptive pulse width adjusting method thereof and electronic equipment | |
KR100219969B1 (en) | 3-level d/a converter and method thereof | |
US6084467A (en) | Analog amplifier clipping circuit | |
JPH01202930A (en) | Signal integrator | |
US6803798B1 (en) | Output control apparatus of pulse width modulator | |
US7136003B1 (en) | Clockless pulse shaping circuit for controlling a power amplified output | |
US20050134249A1 (en) | Circuit arrangement for regulating the duty cycle of electrical signal | |
EP1243072A1 (en) | Switched capacitor filter with integrated voltage multiplier | |
US5220306A (en) | Digital signal comparator for comparing n-bit binary signals | |
US6784710B2 (en) | Multistage pulse width modulator | |
US10298247B1 (en) | Modulators |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AS | Assignment |
Owner name: SONIC INNOVATIONS, INC., UTAH Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:WAYNE, DAVID A.;REEL/FRAME:009587/0936 Effective date: 19980901 Owner name: SONIC INNOVATIONS, INC., UTAH Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:BLYTH, TREVOR;REEL/FRAME:009587/0926 Effective date: 19980629 Owner name: SONIC INNOVATIONS, INC., UTAH Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:NISE, BENJAMIN E.;REEL/FRAME:009587/0934 Effective date: 19981030 |
|
STCF | Information on status: patent grant |
Free format text: PATENTED CASE |
|
AS | Assignment |
Owner name: SILICON VALLEY BANK, CALIFORNIA Free format text: SECURITY INTEREST;ASSIGNOR:SONIC INNOVATIONS, INC.;REEL/FRAME:010485/0439 Effective date: 19991222 |
|
CC | Certificate of correction | ||
FEPP | Fee payment procedure |
Free format text: PAT HOLDER NO LONGER CLAIMS SMALL ENTITY STATUS, ENTITY STATUS SET TO UNDISCOUNTED (ORIGINAL EVENT CODE: STOL); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY |
|
REFU | Refund |
Free format text: REFUND - SURCHARGE, PETITION TO ACCEPT PYMT AFTER EXP, UNINTENTIONAL (ORIGINAL EVENT CODE: R2551); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY |
|
FPAY | Fee payment |
Year of fee payment: 4 |
|
SULP | Surcharge for late payment | ||
FEPP | Fee payment procedure |
Free format text: PAYER NUMBER DE-ASSIGNED (ORIGINAL EVENT CODE: RMPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY Free format text: PAYOR NUMBER ASSIGNED (ORIGINAL EVENT CODE: ASPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY |
|
FPAY | Fee payment |
Year of fee payment: 8 |
|
AS | Assignment |
Owner name: SONIC INNOVATIONS, INC., UTAH Free format text: RELEASE;ASSIGNOR:SILICON VALLEY BANK;REEL/FRAME:025715/0466 Effective date: 20101213 |
|
FPAY | Fee payment |
Year of fee payment: 12 |